AMBER Archive (2004)

Subject: Re: AMBER: GO lie Amber simulation

From: David A. Case (case_at_scripps.edu)
Date: Sat Nov 27 2004 - 17:37:45 CST

On Sat, Nov 27, 2004, Alena Shmygelska wrote:
>
> I am trying to use Amber 7 with GO like model (not using Amber force
> field) I found that when NPHB - number of distinct 10-12 hydrogen bond
> pair types is larger than 500, Amber does not run...

>
> -------------------------------------------------------
> Amber 6 SANDER_CLASSIC Scripps/UCSF 1999
> -------------------------------------------------------
>
> 1. RESOURCE USE:
>
> NATOM = 37 NTYPES = 37 NBONH = 0 MBONA = 36
> NTHETH = 0 MTHETA = 35 NPHIH = 0 MPHIA = 34
> NHPARM = 0 NPARM = 0 NNB = 106 NRES = 37
> NBONA = 36 NTHETA = 35 NPHIA = 34 NUMBND = 36
> NUMANG = 35 NPTRA = 68 NATYP = 37 NPHB = 550
> IFBOX = 0 NMXRS = 1 IFCAP = 0

I am confused: all Amber code that I can see (including Amber 6
sander_classic) checks for the value of NPHB, and aborts the job if it is
too big. This happens at around line 120 of rdparm.f, or search for "NPHB".
You need to figure out why this check is not being performed. If you have
modified the code, you may have a debugging chore on your hands -- be sure
that the ASOL, BSOL & HBCUT arrays in parms.h are at least as big as the
value of NPHB you are using.

...dac

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu